The lab’s robotics division has introduced a new class of collaborative robots — or “cobots” — designed to safely and efficiently work alongside humans in industrial and laboratory settings. Powered by real-time motion prediction and adaptive force control, these robots can anticipate human movement, prevent collisions, and assist with precision tasks. Unlike traditional automation systems that require isolation from human workers, the cobots are built for cooperation. They can dynamically adjust their roles depending on the worker’s pace and the nature of the task, creating a more fluid and efficient workflow. The research demonstrates a promising vision for the future […]
In a major leap toward efficient computing, the lab has developed a new class of edge AI models capable of processing data locally on lightweight devices. Unlike traditional AI systems that rely on cloud-based servers, these models run directly on mobile phones, drones, and embedded sensors — reducing latency, bandwidth use, and energy consumption. This advancement allows AI to function in environments with limited connectivity, from disaster zones to remote agricultural areas. The models can detect patterns, make decisions, and adapt in real time, enabling faster and more autonomous operations. The implications reach far beyond convenience. By decentralizing intelligence, the […]
Researchers at the lab have unveiled a powerful new framework that aims to make artificial intelligence more transparent and accountable. Traditional deep learning systems often operate as “black boxes,” producing highly accurate results without revealing how those results were obtained. The team’s explainable AI (XAI) framework directly addresses this challenge by translating complex model behavior into human-readable insights. The system provides visual heatmaps, textual justifications, and confidence measures for every prediction, helping users understand both the logic and the limitations of AI-driven decisions. Early trials show that this interpretability does not come at the cost of accuracy — in fact, […]